This study investigated the feasibility of using ChatGPT (GPT-4) to create MARC 21 bibliographic records following RDA: Resource Description and Access. This study provided ChatGPT with data from information sources and evaluated its performance in referencing proper records, accurately interpreting information, and correctly applying RDA. The findings demonstrated that although ChatGPT was promising as an assisting tool for human catalogers, it struggled with complex bibliographic patterns and nuanced cataloging rules. Notable errors were observed in various fields, such as primary and additional authorized access points, title statements, edition statements, publication details, and series statements, highlighting areas for improvement.